fix: unescaped dollar signs

llm_web_kit/input/datajson.py

@@ -221,6 +221,7 @@ def __process_nested_list(self, items, list_attribute, indent_level=0, exclude_i
                     continue  # 如果不是dict也不是list，跳过该项
 
             item_text = item.get('c', '')
+            item_text = self.__escape_md_special_chars(item_text)
 
             # 创建列表项行
             item_line = f'{indent}{list_prefix} {item_text}'
@@ -320,6 +321,7 @@ def __content_lst_node_2_md(self, content_lst_node: dict, exclude_inline_types:
             if not title_content:
                 return ''
             level = content_lst_node['content']['level']
+            title_content = self.__escape_md_special_chars(title_content)
             md_title_level = '#' * int(level)
             md_title = f'{md_title_level} {title_content}'
             return md_title

tests/llm_web_kit/extractor/test_extractor_chain.py

@@ -208,7 +208,9 @@ def test_mathlab_html_to_md(self):
         self.assertEqual(result['track_id'], 'mathlab_code')
         md_content = result.get_content_list().to_nlp_md()
         self.assertIn('### Use Integers for Index Variables', md_content)
-        self.assertIn('### Limit Use of `assert` Statements', md_content)
+        self.assertIn('### Limit Use of', md_content)
+        self.assertIn('assert', md_content)
+        self.assertIn('Statements', md_content)
 
     def test_list_to_md(self):
         """测试第三个数据：这个数据会丢失一些文本信息."""
@@ -230,7 +232,9 @@ def test_code_mix_in_list(self):
         input_data = DataJson(test_data)
         result = chain.extract(input_data)
         md_content = result.get_content_list().to_nlp_md()
-        self.assertIn('The descendant of `StandardizerActionRunner` interface has to provide', md_content)
+        self.assertIn('The descendant of ', md_content)
+        self.assertIn('StandardizerActionRunner', md_content)
+        self.assertIn('interface has to provide', md_content)
 
     def test_code_pre_mixed(self):
         chain = ExtractSimpleFactory.create(self.config)